Exploring edX data during 2012 to 2013 by Kan-Hua Lee
Basic information of the features
Number of rows and columns of the data
## [1] 641138 26Number of unique users
## [1] 641138 30Histograms of number of certificates:
##
## 0 1 2 3 4 5
## 613463 23966 2974 609 79 47Univariate Plots Section
Registered users of each course
Certified users of each course
pass.rate of each course
Histogram of age among registered users
plot the histogram of all users vs. LOE_DI, with NA and blank (“”) filtered.
Histogram of gender among registered users
Plot the distribution of nevents
## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
Histogram of access.rate that certified==0
## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
Histogram of access.rate that certified==1
## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
Univariate Analysis
What is the structure of your dataset?
This data has *** column and *** rows. Each row is the data for a user taking a course.
What is/are the main feature(s) of interest in your dataset?
We mainly interested in the backgrounds and involvement of the students that attended and earned certification of each course. The features in this data set that we will focus on are: registered, certified, LoE_DI, YoB, gender, nevents and ndays_act.
What other features in the dataset do you think will help support your investigation into your feature(s) of interest?
Did you create any new variables from existing variables in the dataset?
The following new variables are created in the original dataframe edxdata:
age: the age of the user when taking the course. It is calculated by 2013-YOB.access.period: Day difference betweenlast_event_DIandstart_time_DI.access.rate:ndays_actdivided byaccess.period. This variable essentially measures the how often a users accesses the course.
Also, we grouped the raw data by the following a number of features and created new variables for each data set for each new data sets:
users
The data frame grouped and summarised the data by each user.
course_taken: number of courses viewed.total_registered: number of courses registered.total_explored: number of courses explored.user.certificates: number of courses certified.
course_id
We grouped and summarised the data by course_id. The following new variables are created:
passed_num: total certified users of the course.explored_num: total users who explored the course.registered_num: total users who registered the course.total_nforum_posts: total number of posts in the course forum.pass.rate: the number of certificated users divided by the number of registered usershangon.rate: the number of explored users divided by the number of registered users
course
Some of the course was offered more than one times durin 2012 and 2013. Therefore, another data frame is generated by grouping and summaring the raw data by course_code.
t_certified: number of total certified users.t_viewed: number of users who viewed the course.t_explored: number of users who explored the course.t_registered: number fo users who registered the course.
Of the features you investigated, were there any unusual distributions? Did you perform any operations on the data to tidy, adjust, or change the form of the data? If so, why did you do this?
A few preprocessings of the raw data were performed, as listed below
- Transform the feature
LoE_DIinto levels. - Transform the features
certified,exploredandviewedinto logical type.
Bivariate Plots Section
Histogram of certified users by LoE_DI 
Histogram of certified users by gender 
Distrubution of nevents of certified users only
## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
Boxplot of nevents against course 
Boxplot of ndays_act against course 
Boxplot of access.periods of each course 
Boxplot of access.periods of each course 
Analyze average login events of certified users
## Source: local data frame [13 x 3]
##
## course_code mean_nevents med_nevents
## 1 CB22x 2752.8906 2292.0
## 2 CS50x 284.1166 236.5
## 3 ER22x 1387.1113 1338.5
## 4 PH207x 6144.8561 5465.0
## 5 PH278x 1739.2003 1496.0
## 6 6.002x 5353.4035 4481.5
## 7 2.01x 5312.8340 4826.0
## 8 14.73x 4797.0168 4508.5
## 9 3.091x 7269.2455 6377.5
## 10 6.00x 7953.6741 7227.5
## 11 7.00x 5942.5189 5548.0
## 12 8.02x 9678.3005 9054.5
## 13 8.MReV 7108.4949 6568.0The distribution LoE_DI of certified users 
Distribution of LOE_DI of all users 
ndays_act vs. nevents among the users who has explored more than half of the course: 
access.rate vs. grade 
Bivariate Analysis
Talk about some of the relationships you observed in this part of the investigation. How did the feature(s) of interest vary with other features in the dataset?
Did you observe any interesting relationships between the other features (not the main feature(s) of interest)?
What was the strongest relationship you found?
Multivariate Plots Section
Plot the distribution of nevents
## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
Plot the distribution of nevents of certified users only
## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.
Plot the distribution of access.rate of certified users:
## stat_bin: binwidth defaulted to range/30. Use 'binwidth = x' to adjust this.